The “Mining-Led Growth” in Bourbon Mexico, the Role of the
State and the Economic Cost of Independence1
Rafael Dobado and Gustavo A. Marrero
Universidad Complutense de Madrid
February 2005 (very preliminary version)
Introduction
The Spanish empire in America has always had, and it still has, a
notoriously bad press. It is not easy to find many defenders of the Spanish
colonial history among specialists or ordinary people. In fact, the former as
well as the latter, if asked, would probably pronounce the unanimous
opinion that many of the contemporary economic problems plaguing Latin
America have their roots in the “colonial heritage”. Central to this
standpoint is the view that the colonial State was a factor of economic
backwardness. In this paper we do not examine the general implications of
this popular claim. On the contrary, we focus on the role of the State in the
economic growth of Bourbon Mexico (New Spain) and on the economic cost
of Independence.
In the early 18th century, “a great shift” [Klein (1995)] caused New
Spain to substitute for Peru as the jewel of the Spanish Crown in America.
By 1800, after decades of growth, the remittances of fiscal surplus from
New Spain to the Peninsula and other Spanish colonies in the Caribbean
reached a historical record and became the cornerstone of the imperial
finances [Marichal y Souto (1994)]. In spite of a “pessimist” view about the
economic performance of colonial Mexico in terms of modernization and
convergence with the US [Coatsworth (1990 and 2003)], most indicators
show that during the century preceding the start of the Insurgencia (1810)
the GDP and the population of the Viceroyalty grew. Maddison (2001) has
estimated a per capita GDP growth of one third.
1
Short, somewhat informal, and slightly modified, English version of a previous
article [Dobado and Marrero (2001)]. Please do not to quote without permission by
the authors. Any comment will be welcome.
1
We find that changes in the role of the royal monopoly on mercury
were a decisive factor in the growth of the mining sector that led the
expansion of the rest of the economy. Taxation on an expanding mining-led
economy proved to be a more efficient, and politically acceptable, way to
increase fiscal revenue than maximizing monopoly income from mercury.
Other
measures
(institutional
creativity,
tax
exemptions,
technical
education, etc.) reinforced the positive consequences of that important
change –from strict mercantilism to limited proto-liberalism- in the behavior
of the State. The colonial State, aiming at increasing its fiscal revenue in
New Spain, promoted genuine economic (pre-industrial) growth through, in
particular, its policy on the mining sector.
The decades after the Independence are unanimously considered a
period of economic and demographic decline [Coatsworth (1990), Maddison
2001, Cárdenas (1997 and 2003) and Salvucci (1997)]. The divergence of
Mexico with respect to the Atlantic economies was especially important in
this period. Why did the economic potential of Independence fail to
materialize? [Bulmer-Thomas (1994)]. If the economic burden of the
colonial relationship with an absolutist State and a relatively stagnant
economy was heavy [Coatsworth (1990)], should not Independence bring
with it a new era of growth? Coatsworth’s answer emphasizes the cost of
“gaining
the
Independence”,
that
was
dramatically
high
in
Mexico
(devastating wars and subsequent human and physical capital destructions,
disruption of economic networks, fiscal crisis, etc.), while he regards the
Independence as having benefits only. However, it might be interesting to
explore the possibility that the colonial relationship provided benefits in
some particular, although not irrelevant, economic domains as well. Some
benefits has been already suggested by Bulmer-Thomas (1994). Others
have probably been overlooked by the specialized literature.
The results of our tentative estimation of the economic cost of the
Independence support the idea that Mexico paid a big price for the violent
and long-lasting interruption of the colonial “mining-led growth” from 1810
on.
Therefore, in this paper, three main propositions will be defended:
2
1) A process of pre-industrial
economic
growth
that may
be
described as “mining-lead growth” took place in colonial Mexico
during the 18th century.
2) The Spanish colonial state played a protagonist role in the
“mining-lead growth” through innovative policies and institutional
creativity.
3) Independence itself, and not only because of the costly way in
which it was gained, imposed a very important economic burden
on Mexico.
Silver, mercury and Bourbon reformism
It is widely accepted that silver production grew in Bourbon Mexico at
an unprecedented rate during the 18th century.2 Even though there was no
lack of short and medium term crisis, the long time span of the mining
growth is nonetheless remarkable –see Figure 1. The average yearly rate of
growth between 1690 and 1809 was 1.3%.3 By 1800, New Spain’s silver
production represented almost two thirds of the world output.4 In the
postcolonial period, Mexico continued to be the main silver producer in the
world during most of the 19th and 20th centuries. However, it never
recovered the share of world production it had held in colonial times. Thus,
as Wright and Czelusta (2002) have observed with regard to the US case,
and confining their conclusion to silver production, “no other country
exploited its geological potential to the same extent.”5
In spite of the extraordinary abundance of silver deposits in
numerous locations within the Mexican geography, the sustained growth of
the New Spain’s mining sector is particularly surprising. It took place in the
context of a backward economy. It was based on a “nonrenewable”
2
New Spain mining sector basically consisted in extracting and refining silver
minerals. Gold production was significant but much lower. Production of other
minerals (quarries excluded) and metals was almost negligible in comparison with
silver [Humboldt (1991)]. However, it satisfied domestic demand [Elhuyar (1825)].
3
This rate is higher than that as estimated by Crafts (1994) for the British industry
during the 18th century (1,1%).
4
Mexico’s share of world silver production was 64.4% in 1781-1800 and 62.4% in
1801-1810 [Schmitz (1979)].
3
resource. Besides, ceteris paribus, the uninterrupted exploitation of a given
underground mineral deposit inevitably leads to the appearance, sooner or
later, of diminishing returns. Random increases in the metal content of the
ores, new discoveries, vertical integration
of
firms or technological
innovations may counterbalance that unavoidable tendency of the mining
sector. In this regard, New Spain’s mining sector was far from being an
exception. But, as we will see later on, the Spanish colonial state could, and
it actually did, contribute repeatedly to the postponement of the moment
when silver production would finally become stagnant. Its contribution
included, but was not limited to, very close versions of two of the elements
that David and Wright (1997) identify in the rise of American mineral
production
between
1870
and
1910:
1)
“an
accommodating
legal
environment”; and 2) “education in mining, minerals, and metallurgy”.6
Assessing the “mining-led growth” and the role played by the State in
its achievement requires some additional information. Two alternative
techniques
permitted
the
refining
of
silver
mineral:
smelting
and
amalgamation. The first one was traditionally used in Europe; the second
one, particularly in its variety known as “beneficio de patio”, was a genuine
metallurgical innovation in response to the specific conditions of silver
mining in New Spain (low-grade ores, massive deposits and high price of
fuel). An important difference between the European method (smelting) and
the “beneficio de patio” was that the latter needs a rare and very expensive
metal: mercury. But, fortunately enough for “mineros” –owners of mining
firms-, and for the Crown, Spain was the uncontested main producer of
mercury in the world. Moreover, Almadén mines, the richer mercury deposit
ever exploited was a Spanish Crown property. Commercialization of
mercury was a King’s monopoly. Mercury was sold to miners through the
network of Reales Cajas (fiscal offices established in every miner center and
other important towns). Their accounts has been reconstructed by TePaske
and Klein (1986) and, in particular, Klein (1998) studied them exhaustively
in order to describe the economic and demographic evolution of the main
Viceroyalties of the Spanish Empire between 1680 and 1809. Thus, the
Spanish Crown controlled the supply of an indispensable input in the silver
5
6
Wright and Czelusta (2002), p. 6.
Wright and Czelusta, 2002, p. 6.
4
production process for which there was not substitute. This control implied
not only a mechanism for assuring the “devotion” of the powerful mining
lobby but also an instrument for reducing fiscal evasion.7 Other things being
equal, it was in the interest of the Spanish Crown that amalgamation were
the technical choice of mining firms.8 Actually, mining firms tended in the
long run to opt for amalgamation, albeit not without medium term
fluctuations,9 in response to changes in mercury supply –see Table 1 and
Figure 2. Not surprisingly, the trend of silver production, whether through
amalgamation or total, closely resembles that of mercury sold to miners by
Cajas Reales –see Figure 3.10 Consumption of mercury was also sensible to
price –see Figure 4.11
7
In every mining center there was set a “correspondido” (mercury/silver technical
ratio) which might eventually fluctuated over time in response to the grade of the
silver ore extracted by mining firms. In accordance with the specific
“correspondido” in force in a given mining center, firms were compelled to
“manifestar” –to present- to the Reales Cajas a certain amount of silver
proportional to the mercury they had previously bought. Depending on
circumstances (climate, altitude, type of minerals, etc.) it usually took from several
weeks to several months to complete the amalgamation process of big quantities of
silver ore. The capital requirements of the process and the scale economies of the
“beneficio de patio” explain the higher concentration of firms in the metallurgical
phase of the productive process than in the mining phase as well as the vertical
integration of both phases into big mining firms, which were only a small part of the
total number. Obviously, however inexpensive the “correspondido” was not a
perfect device and it may be regarded as a cheap byproduct of the fiscal
infrastructure. Notwithstanding, it reduced fiscal evasion to an acceptable level.
8
Unless incurring in important information costs, it was simply impossible to control
in a comparable way the production of silver through smelting.
9
The choice of metallurgical technique had a random component: the type of
mineral extracted form new discoveries. In some cases, it was richer enough to be
smelted.
10
Circa 1800, the naval warfare with Great Britain interrupted the regular
communications between the Peninsula and New Spain. Desperate efforts by the
Spanish Navy to transport mercury across the Atlantic failed to secure a normal
supply. That is why some important mining centers partially substituted smelting
for amalgamation. Besides, as a means of overcoming the crisis, miners obtained
exemptions or postponements of taxes from the fiscal authorities. When they finally
paid, the entry on the official books was dated as of the current year. Both
circumstances help to explain why the decrease in mercury consumption around
1800 was not paralleled by that of silver production.
11
The estimation (OLS) in first differences (variables in logs) of a simple model of
demand in which HGCONSUMPTION = f(HGSTOCK, HGPRICE) yields the following
results:
HGCONSUMPTIONt = -0.004 + 0.703*HGSTOCKt – 1.151*HGPRICEt-2 + εt
(-0.110) (8.010)
(-1.835)
Adjusted R-squared = 0.425
Durbin-Watson statistic = 2.632
5
It was the growing tendency of mercury production in Almaden mines
–a government-managed and owned firm- what made it possible to
increase supply and consumption of mercury, which ultimately permitted
the expansion of silver production. But this expansion would not have been
so great if the increase in mercury supply had not been accompanied by
reductions in prices.
Contrary to what most specialists claim, Bourbon reformism started in
the first half of the 18th century. In 1717, the Superintendencia General de
Azogues was created with the aim of centralizing the decisions about
mercury in the Peninsula as well as in New Spain and of giving more power
to the officers of Indias -“Ministry of Colonies”. In 1723, taxes on silver
production were reduced substantially: 50% (from one fifth to one tenth).12
By the middle of 1720’s, the trend of mercury production in Almaden curved
irreversibly upwards while its volatility and that of exports were reduced.
After 1739, transportation of mercury was exempted from the general, too
restrictive, rules governing the commerce between the Peninsula and the
Spanish America whit the subsequent increase of flexibility in the supply of
mercury to silver producers. However, it was in the early 1740’s when the
trend of mercury production reached an historical record high and a
minimum threshold for the rest of the century. The reason is twofold: 1) a
series of significant increases in the budget of Almaden –see Figure 5; and
2) first steps along a process of technological innovation that culminated in
the introduction of steam power into Spain, for the first time in civilian
applications, in the late 1780’s and of an original system for mining vertical
deposits of mineral at the end of the century. Neither of these factors
converted Almaden into an example of productive efficiency. But they
permitted a substantial increase in the volume and the regularity of mercury
supply to New Spain mining firms.
In spite of these important changes in the policy on silver and
mercury, production in New Spain decreased from the early 1750’s, when
record levels were reached, to the late 1760’s. It was then when an almost
revolutionary measure was adopted in order to invert the current trend of
12
This measure had been previously adopted in 1710 in Zacatecas, the main
mining center in New Spain. According to Elhuyar (1825), fiscal revenue not only
6
silver production. What made it possible was a deep transformation in the
mentality of colonial officers at both shores of the Atlantic. Until then, the
royal monopoly on mercury was considered, not without good reasons, an
essential component of the Crown’s revenues in New Spain. Afterwards, it
was seen as an instrument to propel the “mining-led growth” and its fiscal
implications [Dobado (2002)].
“Mining-led growth” is simply a model of pre-industrial growth in
which mining played a key role as the driving force of the expansion of the
whole –backward- economy and of its mercantile, more dynamic, sector in
particular.13 The concept is an adaptation to nowadays economic jargon of
information
and
opinions
taken
directly
from
the
most
relevant
contemporary observers (Humboldt14 and Elhuyar15, i. e.) and from a huge
amount of historical documents preserved in archives of Spain and Mexico.
In open contrast with the “resources curse” hypothesis, the concept
recognizes the singular importance of the backward linkages of mining in
New Spain economy and the impossibility that other sector could replace it.
Given the conditions of New Spain economy, among which internal
transportation
problems16
and
distance
to
main
world
markets
are
extremely important and should be taken into account, silver was probably
the only possible exportable product of the mining sector because of its high
value per unit of weight. For most regions in the highlands of the central
did not decrease but increased. Colonial government deserves some credit for
discovering an early ad hoc version of the Laffer curve.
13
The important subsistence sector –mostly associated with indigenous peoples- of
New Spain economy stood aside the “mining-led growth”.
14
“El laboreo de las minas lejos de ser contrario a la agricultura, ha favorecido los
desmontes en las regiones más desiertas.” (Humboldt, 1991, p. 566).
15
“En comparación a las ventajas que (...) proporciona la minería, debe estimarse
en poco el valor de dichos preciosos metales. ¿Qué son en efecto veinte ni veinte y
siete millones [de pesos] a que ha llegado la acuñación de este reino, respecto de
lo que importa su tráfico interior, debido en la mayor parte al impulso de las minas?
¿Qué es su alucinante (...) riqueza, respecto de la que presentan los floridos valles,
vegas y dilatadas llanuras cultivadas, las selvas y potreros cubiertas de ganados,
las numerosas poblaciones esparcidas en la vasta extensión de su suelo, las
manufacturas, artes y oficios que en ellas se ejercen, y en fin el considerable
número de habitantes que cuenta en su seno y que igualmente son deudores de su
floreciente estado al influjo y trascendencia del propio ramo que todo lo ha
vivificado y sigue sosteniendo y fomentando?” (Elhuyar, 1825, p. 16).
16
In reference to 19th century, Coatsworth highlights that, in Mexico, “the
geography conspires against the economy” (Coatsworth, 1984. p. 19). Elhuyar
(1825) and Ward (1981) also emphasized the negative influence of Mexican
geography on inter-regional trade and specialization.
7
plateau and in the remote and arid northern steppes, there was simply no
economic alternative (agriculture or manufactures) to silver production that
were capable of entering into the world market. In fact, modern Mexico has
been shaped by mining expansion across its rather hostile geography –see
Map 1. Mining contributed as no other economic activity to regional
integration and to urban expansion in so wide a country as no other
economic activity ever did. From the viewpoint of the endowment of natural
resources in Mexico and its related economic choices in the 18th century, the
question of whether or not it was a Spanish colony bcomese doubtfully
relevant. In other words, colonial Mexico silver exports did not only respond
to the Spanish Crown will but to economic rationality. In the postindependence period silver continued being one of the main economic
activities and the most important exportable product in Mexico.
Besides, contrary to what Acemoglu, Johnson and Robinson (2001)
seem to believe, silver mining was far from being a typical case of
extractive institution. Their hypothesis about the “reversal of fortune” and
the role of European colonialism in explaining the striking differences in
world distribution of income might or not be true, but it does not apply to
New Spain’s mining sector. First of all, silver production was exclusively
based on free labor. Those numerous mining centers located away from the
valley of Mexico could only develop trough migration of free workers. As in
most mining centers there were not dense societies from which labor could
be easily extracted, silver miners had bargaining power and were very well
paid. According to Humboldt, the Mexican miner was “the best paid among
all miners”.17 Although some owners of mines were incredibly rich and
prone to sumptuary expenses, most mining firms were of medium, small or
minimum size. Ownership of mines was not very unequally distributed. The
real picture of mineral wealth in New Spain differs from that of extreme
inequality suggested by Engerman and Sokollof (2002) for Latin America.
It also probably unknown that the institutional framework of mining
in New Spain was very close to the type referred as “private property” by
Acemoglu,
Johnson
and
Robinson
(2002).
In
particular
after
the
promulgation in 1783 of the Reales Ordenanzas –mining code-, access to
property rights of mineral deposits was easy and cheap and its protection
8
was guaranteed. In 1778, owners of mining firms in all regions were
authorized to be organized in a peculiar representative institution called
Real Tribunal General de la Minería. The mining lobby was thus given an
instrument to increase its influence on the economic decision-making
process. By a royal order, the Real Tribunal was financed by menas of a
transfer from the taxes collected by the State on silver production and
minting. The Real Tribunal permitted a more efficient defense of mining
firms interests, even tough it was not fairly equitable and was managed
with criteria that may be criticized. It was also in charge of the Escuela de
Minas (Mining School) established in the capital in 1792. Humbodlt (1991)
praised this scientific and educational center. Quoting this qualified
contemporary observer, Klein (1995) reckons that the fiscal pressure
exerted on mining was lower than the European average.18 General or
particular fiscal exemptions on the purchase of certain taxed inputs or with
a view to promoting new discoveries or re-exploiting old abandoned
deposits were common during the last decades of the 18th century.
Therefore, there was not any predator colonial State confiscating the results
obtained by the individuals from their productive efforts in New Spain’s
mining sector. Not surprisingly, for some time, a variety of documents
reflects a lasting “honey moon” between mining firms and the Spanish
Crown. The institutional creativity and adaptive behavior of the colonial
State is not independent of this peculiar estate of affairs.
Among the main changes that may be included within this “second
wave” of reforms initiated in response to the stagnation of silver production
by the middle of the 18th century is the above-mentioned reorientation of
the goals set by the royal monopoly on mercury. Two consecutive decreases
in mercury prices (1767 and 1776) halved the cost of this strategic input for
mining firms in New Spain –see Figure 4. The positive effects of this
measure did not take long to appear –see Table 1. Lower prices of mercury,
combined with an increasing supply, made possible the rise in rates of
growth of silver production as shown in Table 1. It is interesting to note that
the shift in the traditional behavior of the royal monopoly on mercury was
17
Humboldt, 1991, p. 370.
The estimation of the fiscal pressure with original data leads to the same
conclusion.
18
9
not aimed at increasing its income but to maximize the total fiscal revenue
in New Spain –see Table 2. In fact, the decrease in prices was not followed
by an increase in the royal monopoly income, as it might be expected given
its market power and the elasticity of demand for mercury. Until 1777,
benefit per unit of mercury sold in New Spain was fabulous. Afterwards,
price of mercury was not only lower than in Spain and the international
market but also it ceased to provide significant profits. However, the losses
associated to the “political prices” charged by the monopoly were
counterbalanced in excess by the substantial increases in taxes on silver
production and minting and in total taxes collected –see Table 2.
For the Crown, the essence of the “mining-led growth” may be
described by means of the following fundamental economic relationship:
∆ MERCURY CONSUMPTION ⇒ ∆ PRODUCTION SILVER ⇒
⇒ ∆ ECONOMIC ACTIVITY ⇒ ∆ TAX COLLECTION ⇒
⇒ ∆ FISCAL SURPLUS
The
co-evolution
of
mercury
consumption
with
fiscal
surplus
remittances from New Spain to the Peninsula and other colonies in the
Caribbean is simply amazing –see Figure 6. Long term relationship (17201800) between mercury consumption and fiscal surplus remittances may be
captured by a cointegration equation [Dobado (2001)]. Besides, an
additional short term relationship may be captured by an error-correction
model [Dobado (2001)].
Explaining the active mining policy adopted by the Spanish Crown in
New Spain does not need to assume philanthropic motivations. On the
contrary, two simple assumptions will suffice: 1) rational pursuit of selfinterest; and 2) capacity of improving perceptions of the economic process.
Efficient governmental actions towards New Spain’s mining sector were
encouraged by the open political and military conflict between Spain and
Great Britain existing all along the 18th century in the international arena. In
spite of being lost by Spain, this conflict may be interpreted as a sort of
selective pressure upon the efficiency of the Spanish State. In other words,
the imperial rivalry –various costly wars in different oceans, islands and
10
continents included- between both monarchies would have finished earlier
and with bigger loses for the Spanish Crown if reforms –limited as they
were- in the State would not have been introduced. The renovation of
political and bureaucratic elites resulting from the change of dynasty also
had favorable consequences on the management of the economic affairs by
the Spanish State.19 Increasing the fiscal revenue was necessary to sustain
the war efforts. Economic prosperity appeared as a less conflictive way to
augment the fiscal base of the Empire than its alternative –increasing the
fiscal pressure. Coherent with the political and military goals of the Bourbon
State,
the
economic
policy
experienced
a
gradual
change
from
interventionist mercantilism towards some form of limited proto-liberalism
with a bigger potential to promote pre-industrial growth.
Proposition 1: A process of pre-industrial economic growth that may be
described as “mining-lead growth” took place in colonial Mexico during the
18th century.
Our argument in favor of this proposition consists of showing the
existence of a significant relationship between silver production and tax
collection. Our analysis
distinguishes
between
short
and
long
term
relationships. The economic assumption underlying this argument is that
tax collection paralleled economic activity. As the flexibility of the fiscal
structure of colonial Mexico with respect to economic conjuncture may be
questioned, we have worked on the series constructed with data provided
by TePaske and Klein (1986) in order to eliminate all entries on the fiscal
accounts that might be uncorrelated with the level of activity of New Spain’s
economy. Thus, we are taking into account criticism of an incautious
confidence in fiscal sources as proxies of economic activity [Brading (1985),
Pérez Herrero (1996) and Klein (1994)]. The result is the series INGORD1:
ordinary receipts of the Crown in the three main Cajas Reales (1711-1800).
To prevent the bias that mining taxes and mercury sales might introduce in
the analysis, we have deducted them from INGORD1. The new series
(INGORD2) includes all receipts except those depending directly from the
19
Bourbons substituted for Austrians at the beginning of the 18th century.
11
mining sector. An additional series (INGORD3) has also been calculated to
test the robustness of the results regarding different definitions of fiscal
receipts. INGORD3 excludes the tributo indígena (a capitation tax paid by
every native adult male). This series is very likely to proxy those economic
activities that took place in the market. The coevolution of silver production
with the three series of fiscal receipts is depicted in Figure 7. The results of
the cointegration analysis are shown below. In all three cases, we reject H0
at the 1% level of signification. Therefore we accept the hypothesis of
cointegration. Results are shown below.20
Cointegration equation between silver production and ordinary
receipts of the Crown in New Spain, 1711-1800.
α
β
ρ
Tρ
INGORD1
-11.26
1.69
-0.50
-5.15
INGORD2
-7.43
1.42
-0.68
-6.78
INGORD3
-9.52
1.52
-0.62
-6.13
A significant and stable long term relationship between silver
production and ordinary receipts is not rejected at 5% level of significance.
β is interpreted as an estimation of the elasticity between silver production
and ordinary receipts throughout the long term equilibrium path. In all
cases, it is higher than 1 which seems to suggest a multiplier effect of the
mining sector on the non-mining receipts and hence on the level of activity
of non-mining sector.
The cointegration analysis supports the hypothesis about the “miningled growth” in colonial Mexico as a long term relationship. To check whether
or not this relationship is reinforced by an additional short term relationship,
we have estimated an error-correction models between silver production
20
In all three cases we contrast H0: ρ=0, against H1: ρ≠0, in the model
ut = ρ ut -1 +
k
∑ π ∇u
i =1
i
t -i
+ w t , where ut is the residual (OLS) of the cointegration
equation ln(Yt ) = α + β ln(Xt ) + ut , where Yt is INGORD1, INGORD2 or INGORD3
and Xt is silver production. k is taken big enough as to allow for wt being white
noise. The statistic of contrast is Tρ = ρ / σ ρ , which is compared with critical values
in Engle y Yoo (1987).
12
and the three definitions of ordinary fiscal receipts. The results are shown
below. 21
Error-correction models of silver production and ordinary receipts of
the Crown in New Spain, 1711-1800.
φ
δ x0
δ y1
DW
R2
INGORD1
INGORD2
INGORD3
0.43
0.53
--
-0.38
---
-0.24
-0.58
-0.49
0.33
0.29
0.26
2.11
2.08
2.23
The estimation of the ECM models implies that a change in ordinary
receipts responds to contemporary changes in silver production and to past
changes in receipts. Additionally, deviations of receipts from their long term
equilibrium relationship with silver have a negative effect on their growth
rates.
Our interpretation of these results is that mining led the growth of the
economic activities from which ordinary taxes were collected. These
activities belonged to the mercantile, most dynamic, sector of New Spain’s
economy. In sum, the “mining-led growth” hypothesis is the least
improbable explanation of the narrow relationship observed between silver
production and fiscal income.
Proposition 2: Spanish State –metropolitan and colonial- had a protagonist
role in the “mining-lead growth" through innovative policies and institutional
creativity.
This proposition is based on abundant historical evidence shown in
the second section of this article. Besides, we offer statitical support in favor
of this proposition. To this goal, we specify and estimate a dynamic model
that relates silver production (AGPROD) to which policy variables –price of
21
All parameters are significant to the 5% level. For every ordinary receipts
the
ECM
model
is
as
follows:
definition
(Yt,),
py
∇ lnYt =
∑ δyi ∇ lnYt -i +
i =1
px
∑δ
j =0
xj
∇ lnXt -j + φ ut −1 + ε t , where Xt is silver production and
ut −1 shows the extent to which the relationship between Yt and Xt differs in the
13
mercury (HGPRICE) and budget of Almaden mines (ALMADENBUDGET)- and
the
stock
of
mercury
in
New
Spain
(HGSTOCK).
These
variables
(ALMADENBUDGET, HGPRICE and HGSTOCK) try to capture the effects of
the important changes in the State’s behavior during the 18th century on
silver production (AGPROD). As the mining sector interacted with the
agricultural sector, we have also included the price of corn (CORNPRICE) in
the specification of the model. Corn was the main agricultural input for
silver mining. We agree with Salvucci (1994) on the idea that unproductive
Mexican agriculture constituted an important impediment to growth.
Our first step has consisted in contrasting the hypothesis of
cointegration between AGPROD and the explanatory variables. We find that
cointegration is not rejected at
5% level of significance.22 We have then
proceeded to analyze the short term dynamic of the model by specifying
and estimating the following error-correction model:
∇ln(AGPROD) = α + β1∇ln(HGSTOCK (-1)) + β2∇ln(ALMADENBUDGET (-2))
+ β3∇ln(HGPRICE(-1) + β4∇ln(CORNPRICE) + δ RESIDUALSCOINT(-1) +ωt,
The time span of our different specifications –see below- respond to
hypothetically influential facts. The period 1720-1805 is justified by the
availability of reliable data. In 1732, the Mint of Mexico, to which silver the
legally produced was sent for minting, started to be administrated directly
by the Crown. In 1753, mines of Almaden’s budget was substantially
increased by the Ministry of Finances and the beginning of a tendency
towards higher corn prices becomes perceptible. Dummies capture the
influence of abnormal and specially influential values that might bias the
estimation of the relevant parameters of the equation. The results of the
estimation are shown below:
previous year from the long term equilibrium. R 2 is the adjusted coefficient of
determination; DW is the Durbin-Watson statistic.
22
The cointegration equation is:
ln(AGPROD) = 14,5 + 0,03ln(HGSTOCK)
0,05ln(HGPRICE) - 0,04ln(CORNPRICE)
14
+
0,15ln(ALMADENBUDGET)
-
Error-correction models of AGPROD and explanatory variables.
1
2
3
1720-1805
1720-1805
1732-1805
0.06 (*)
0.07
HGSTOCK (-1)
0.05 (**)
(0.03)
(0.03)
(0.03)
0.09
0.12
ALMADENBUDGET(-2)
0.07 (*)
(0.03)
(0.03)
(0.04)
-0.04
-0.04
HGPRICE (-1)
-0.04 (*)
(0.02)
(0.01)
(0.01)
-0.06 (*)
-0.06 (*)
CORNPRICE
-0.05 (**)
(0.03)
(0.03)
(0.03)
RESIDUALSCOINT (-1)
-0.44
-0.31
-0.29
(0.09)
(0.08)
(0.09)
Dummy 1736-37
-0.29
0.30
(0.07)
(0.06)
Dummy 1772
-0.25
0.25
(0.09)
(0.09)
Dummy 1774
--0.33
-0.34
(0.09)
(0.09)
R2
0.33
0.54
0.61
DW
2.28
2.22
2.19
P-valor (residuals
0.66
0.18
0.32
normality test)
P-valor (Q-LjungBox)
0.94
0.89
0.49
4
1753-1805
0.05 (*)
(0.02)
0.47
(0.14)
-0.04
(0.01)
-0.14
(0.04)
-0.33
(0.11)
-0.30
(0.09)
-0.35
(0.09)
0.67
2.38
0.31
0.27
Coefficients are significant at 1% level unless otherwise indicated;
(**) significant at 10%; (*) idem at 5%; standard deviations in brackets.
Our interpretation of these results is that colonial State innovative
policies had positive consequences on silver production while backward the
agricultural sector, whose tendency towards higher prices was accompanied
by recurrent crisis of subsistence
(1770-1774, 1780-1781, 1785-1787,
etc.), became an important factor that limited the potential growth of silver
production. The State’s passivity with respect to agriculture openly
contrasts with its active policies on mining. In spite of the restrictions that
geography impose on Mexican agricultural productivity, institutional change
in agriculture had an important role to play that was never tried.
Proposition 3: Independence itself, and not only because of the costly way
in which it was gained, had a very important economic cost for Mexico.
15
Estimates of Mexican GDP, per capita or total,
during the post-
colonial period are unanimously somber [Coatsworth (2003), Maddison
(2001) and Salvucci (1997)]. By the middle of 19th century Mexican per
capita GDP would be either equal or, more probably, lower than that of circa
1800.
We identify the economic cost of Independence with the interruption
of the “mining-led growth” in Mexican economy. By 1860, silver production
had not still recovered the maximum level reached in the last years of the
Spanish colonial rule in Mexico –see Figure 1. The violence associated with
the Insurgencia (1810) did not only cause an enormous temporary damage
to the mining sector. Extremely important as it was, this damage could
have been overcome sooner or later. What made the interruption of the
“mining-led growth” to have a long-lasting effect in terms of per capita GDP
was the inability of the post-colonial Mexican State to recreate the
conditions in which silver production had flourished before Independence.
State intervention in decisive adverse conjunctures was able to increase the
growth of silver production above its “natural” rate: expanding the supply of
mercury in the 1740’s and lowering its price in the late 1760’s and 1770’s in
particular –see Figure 8.23 The gap between the silver production forecast
for 1809 ant its real trend is especially noticeable. Therefore we believe that
the absence of a “selfish” State, although rational and effective in the
pursuit of its interest, has to be seriously taken into account in order to
explain the crisis of mining after the beginning of the process of
Independence and its very negative consequences on Mexican economic
performance during five decades.
Neither the Mexican state nor the foreign mining firms that started to
operate in Mexico shortly after the Independence proved to be able of
taking advantages of the supposedly great opportunities that the rupture of
the colonial relationship offered. In fact, some British mining ventures failed
to restore the previously buoyant state of some mining centers [Velasco et
al. (1988)].24 In some way, it is ironic that companies from the uncontested
23
The contrast is clear even if the rate of increase of silver production would have
been close to zero from 1809, which is a very implausible hypothesis if the pre1810 colonial State, or something similar, would have continued existing.
24
In January 1824, the share price in London of the British company Real del Monte
was 1,479 sterling pounds, compared with a nominal value of 400 while, by late
16
world economic leader of the moment were not able to get almost any
positive
result
from
their
organizational,
financial
and
technological
superiority over the rest of the world.
Our tentative estimation of the cost of Independence is based on
accepting the existence of a long term stable relationship between silver
production and economic activity in New Spain –see Proposition 1 above. It
is possible then to calculate a conjectural Mexican GDP as a function of
silver production from 1810 to 1860.25 Thus, we compare the evolution of
this Mexican GDP with a counterfactual, which is based on the hypothesis
that
the
“mining-led
growth”
Independence after 1810.
might
have
continued
unaffected
by
Ponzio (1998) has convincingly argued against
the Coatsworth’s pessimist view on the mining sector prospects in the years
before the Insurgencia (1810). Our conjectural GDP (CONJGDP) is based on
the assumption that the share of the mining sector was 10% until 1815 and
then, in response to the profound crisis of silver production, decreased to
5%.26 Thus, we have calculated CONJGDP as a proportion of actual silver
production from 1810 to 1860 -see Figure 9.27 This conjectural path of
Mexican
GDP
is
compared
with
three
counterfactuals:
1)
MINGDP,
“pessimist” (no growth); 2) MAXGDP, “optimist” (yearly average growth
rate of 1.28%);28 and 3)
INTGDP, “intermediate” (yearly average growth
rate of 0.64 %). Except in the improbable case of no growth after 1809,
Independence proves to be very costly in economic terms.29 The most
probable path of Mexican GDP is only occasionally above its potential level.
However, in order to make more realistic our numerical exploration of
the economic cost of Independence we also need to take its benefits into
1848, it fell to 0.63 pounds [Randall (1977)]. This case is probably exceptional but
it illustrates the sad difference between expectations and achievements with regard
to Mexican mining after the Independence.
25
Salvucci (1997) had estimated a GDP of 227,5 pesos circa 1800, which is the
average of available estimations. Mining share was 10%.
26
This assumption –no increase of mining share in GDP in spite of the silver
production recovery since the middle of the 1820’s- is biased the result against our
hypothesis about the casuse of economic stagnation in independent Mexico.
27
It is reassuring that our calculation is very close to those of Salvucci (1997) for
1840 (-1.4%) and of Coatsworth (1990) (7,2%).
28
This rate is obtained as the forecast for 1810-1860 of the univariant model of the
silver production trend (estimated using the Hodrick-Prescott filter) from 1700 to
1809. More details in Dobado and Marrero (2001).
29
In yearly average, MAXGDP is 64.8% higher than CONJPIB. The difference
decreases to 35% for INTGDP.
17
account. Coatsworth (1990) had estimated the cost of the surplus fiscal
remittances and commercial restrictions imposed by Spanish Crown
between 1790 and 1820 at 7.2% of New Spain’s GDP.
We consider this
estimation a genuine cost of Mexico’s colonial relationship with Spain and
therefore a benefit of Independence. We identify the cost of Independence
with the respective differences (%) between CONJGDP and the three
hypothetical paths depicted in Figure 9. Aiming at biasing the results of this
numerical exercise on the cost-benefit analysis of Mexican Independence
against our Proposition 3, we have also examined the hypothesis that the
cost of the colonial relationship were 100% higher than estimated by
Coatsworth (1990).
The results are shown in Table 3 and may be
interpreted as a numerical confirmation of the idea underlying Proposition 3.
It is our guess that the most probable real economic cost of Independence
might amount to some figure between a yearly average of 27.8% of
Mexican –“non-independent”- potential GDP and 50%.
Thus, the main origins of Mexican comparative backwardness does
not date from the 18th century, as it might be deduced from the inadequate
comparison with the exceptional case of the US. Instead, it really dates
from the 19th century, which was the time when most European economies
-Spain and other more comparable countries than the US- started to
experience conspicuous modern economic growth.
In the meantime,
Mexico did not only fail to grow at similar rates but it also needed several
decades to simply recover the GDP reached at the end of the colonial
period.
Mexico’s economic development in the 19th-20th centuries was
not
affected by any supposed “resources curse” or by an insurmountable
“colonial heritage” but by several “lost decades” of growth in both centuries.
18
References
ACEMOGLU, D., JOHNSON, S. y ROBINSON, J. A. (2001), “Reversal of
Fortune: Geography and Institutions in the Making of the Modern World
Income Distribution”, NBER Working Paper, 8460.
BRADING, D. A. (1983), Mineros y comerciantes en el México borbónico,
1763-1810, Fondo de Cultura Económico, México.
(1985), "Facts and Figments in Bourbon Mexico”, Bulletin of Latin American
Research, 4,1, pp. 61-64.
CARDENAS, E. (1997), "A Macroeconomic Interpretation of NineteenthCentury Mexico", HABER, S. (ed.) How Latin America Fell Behind, Stanford
University Press, Stanford, pp. 65-92.
-- (2003), ¿Cuándo se originó el atraso económico de México?, Biblioteca
Nueva-Fundación Ortega y Gasset, Madrid.
COASTWORTH, J. (1984): El impacto económico de los ferrocarriles en el
porfiriato. Crecimiento contra desarrollo, Era, México.
-- (1990), Los orígenes del atraso, Alianza Editorial Mexicana, México.
-- (2003), “Mexico”, MOKYR, J. (ed.), The Oxford encyclopedia of economic
history, vol. 3, pp. 501-507.
CRAFTS, N. (1994), “The Industrial Revolution”, FLOUD y MCCLOSKEY
(eds.), The Economic History of Britain since 1700, Cambridge University
Press, Cambridge, pp. 44-59.
DAVID, P. y WRIGHT, G. (1997), “Increasing Returns and the Genesis of
American Resource Abundance”, Industrial and Corporate Change, 6, 2, pp.
203-245.
DOBADO, R. (1989), El trabajo en la Minas de Almadén, 1750-1855, Tesis
doctoral no publicada, Universidad Complutense, Madrid.
--(1990), “Las Minas de Almadén, el monopolio del azogue y la producción
de plata en Nueva España en el siglo XVIII”, mimeo.
(1997 “Las Minas de Almadén, el monopolio del azogue y la producción de
plata en Nueva España en el siglo XVIII”, SÁNCHEZ, MIRA y DOBADO
(eds.), La savia del Imperio. Tres estudios de economía colonial Ediciones
Universidad de Salamanca, Salamanca, pp. 401-495.
- (2002), “El monopolio estatal del mercurio en Nueva España durante el
siglo XVIII”, Hispanic American Historical Review, 82:4, pp. 685-718.
DOBADO, R. y MARRERO, G. (2001), “Minería, crecimiento económico y
costes de la Independencia en México”, Revista de Historia Económica, XIX,
3, pp. 573-611.
ELHUYAR, F. de (1825), Memoria sobre el influjo de la Minería..., Imprenta
de Amarita, Madrid.
ENGLE, R. F. and GRANGER, C. W. (1987), "Cointegration and error
correction: repreaentation, estimation and testing", Econometrica, 55, 251276.
ENGLE, R. F. and YOO, B. S. (1987), "Forecasting and testing in
cointegrated systems", Journal of Econometrics, 35, 143-159.
ENGERMAN, S. L. and SOKOLOFF, K. L. (2002), “Factor Endowments,
Inequality, and the Paths of Institutional and Economic Development among
New World Economies”, http://www.cid.harvard.edu/Economia/Papers%20
Upcoming%20Meeting/Sokoloff.PDF
HEREDIA, A. (1978), La renta del azogue en Nueva España, 1709-1751,
Escuela de Estudios Hispanoamericanos de Sevilla, Sevilla.
19
HUMBOLDT, A. (1991), Ensayo político sobre el reino de la Nueva España,
Editorial Porrúa, México.
KLEIN, H. (1994), Las finanzas americanas del imperio español, Instituto de
Investigaciones Dr. José María Luis Mora, etc., México.
-- (1995), “The Great Shift: The Rise of Mexico and the decline of Peru in
the Spanish American Colonial Empire, 1860-1809,” Revista de Historia
Económica, XIII, 1.
-- (1998)The American finances of the Spanish Empire, University of New
Mexico Press, Alburquerque.
MADDISON, A. (2001), L'économie mondiale: une perspective millénaire,
OCDE, Paris.
PEREZ HERRERO, P. (1996), “Reformismo borbónico y crecimiento
económico en la Nueva España”, GUIMERA, A. (de.), El reformismo
borbónico, Alianza Universidad, Madrid, pp. 75-107.
PONZIO, C. (1998), “Interpretación económica del último período colonial
mexicano”, Trimestre Económico, Ene-Mar., pp. 99-126.
RANDALL, R. W. (1977), Real del Monte: una empresa minera británica en
México, Fondo de Cultura Económica, México.
SALVUCCI, R. (1994), "Economic Growth and Change in Bourbon Mexico: A
Review Essay", The Americas, 51:2, pp. 219-231.
-- (1997), "Mexican National Income in the Era of Independence, 18001840", HABER, S. (ed.) How Latin America Fell Behind, Stanford University
Press, Stanford, pp. 216-242.
SCHMITZ, C. J. (1979), World Non-Ferrous Metal Production and Prices,
1700-1976, Frank Cass and Co., London.
TEPASKE, J. J. y KLEIN, H. (1986), Ingresos y egresos de la Real Hacienda
en Nueva España, 2 vols., INAH, México.
VELASCO, C. et alii (1988), Estado y Minería en México, 1767-1910, Fondo
de Cultura Económica, México.
WRIGHT, G. and CZELUSTA, J. (2002), “Exorcizing the Resource Curse:
Minerals as a Knowledge Industry, Past and Present”, Working Paper,
Stanford University.
20
Map 1:
21
Figures
Figure 1: Silver production in México, 1690-1860.
30000000
25000000
Pesos
20000000
15000000
10000000
5000000
0
1700
1750
1800
1850
Figure 2: Proportion of silver produced by amalgamation and stock of
mercury in New Spain, 1714-1805.30
40000
80
70
30000
25000
60
20000
15000
50
10000
5000
40
20
30
40
50
60
hgstock
30
70
80
90
amalgamation
Hereinafter, trends are calculated with the Hodrick-Prescott filter.
22
00
%
Castilian quintals
35000
25000
25000000
20000
20000000
15000
15000000
10000
10000000
5000
Pesos
Castilian quintals
Figure 3: Trends of silver production (by amalgamation and total) and
mercury consumption, 1714-1805.
5000000
0
0
20
30
40
50
60
70
80
90
00
hgconsumption
agtotalproduction
agamalgamation
Figure 4: Price and consumption, stock, and production (trends) of mercury,
1714-1805.
90
40000
30000
70
20000
60
10000
50
40
0
20
30
40
50
60
hgprice
hgconsumption
70
80
90
hgstock
hgproduction
23
00
Castilian quintals
Pesos/Castilian quintal
80
Figure 5: Trends of Almaden’s budget and of production of mercury, 17001805.
8000000
25000
20000
15000
4000000
10000
2000000
Castilian quintals
Reales de vellón
6000000
5000
0
1700
0
1720
1740
1760
Almadenbudget
1780
1800
hgprod
Figure 6: Mercury consumption and fiscal surplus remittances, 1723-1797.
16000
10000000
8000000
12000
6000000
10000
8000
4000000
6000
2000000
4000
2000
0
30
40
50
60
hgconsumption
24
70
80
remittances
90
Pesos
Castilian quintals
14000
Figure 7: Trends of silver production and of fiscal receipts, 1711-1800.
Pesos.
25000000
20000000
20000000
15000000
15000000
10000000
10000000
5000000
5000000
0
20
30
40
50
60
agtotalproduction
ingord1
70
80
90
00
ingord2
ingord3
Figure 8: Real trend of silver production and forecasts for selected years,
1690-1860.
17.5
17.0
16.5
16.0
15.5
15.0
1700
1750
1800
forecast 1742-1756
forecast 1767-1781
1850
forecast 1810-1824
agprodtrend 1690-1860
25
Figure 9: Alternative paths of growth of Mexican GDP, 1810-1860.
500
400
Pesos
300
200
100
0
10
15
20
25
30
conjgdp
intgdp
26
35
40
45
maxgdp
mingdp
50
55
60
Tables
Table 1: Yearly (%) rates of growth, 1715-1805.31
1715-1805
1715-1767
Total silver production
1,56
0,97
(New Spain)
Amalgamated silver production
2,04
1,47
(New Spain)
Mercury production
1,83
1,20
(Almaden, Spain)
Stock of mercury
1,61
2,57
(New Spain)
Mercury consumption
1,64
0,80
(New Spain)
1768-1805
2,37
Source: Dobado (1989 and 1997) and Heredia (1978)
1750-1759
1760-1767
1768-1777
1778-1785
1750-1759
1760-1767
1768-1777
1778-1785
Table 2: Yearly average
Panel I: Mercury sales
I
II
Castilian
Pesos
quintals
7.750
640.000
7.300
610.000
10.400
630.000
12.900
505.000
Panel II: Gross Crown’s receipts
IV
V = IV + I
1.210.000
1.850.000
1.235.000
1.845.000
1.495.000
2.125.000
1.780.000
2.285.000
III = II/I
Price
82,6
83,6
60,6
39,1
VI
5.640.116
5.770.435
7.712.095
12.327.517
I: Castilian quintals; II: Pesos; III: Price (pesos/quintal);
IV: Taxes on silver (pesos); V: Pesos; VI: Total taxes
collected in Guanajuato, México and Veracruz (pesos).
Source: TePaske and Klein (1986) and Dobado (1990)
31
Average of first differences of variables in logs.
27
2,83
2,71
0,28
2,83
Table 3: Tentative calculation of economic costs (+) or benefits (-) of
Independence, 18101-1860: Yearly (%) average of GDP.
Panel A
Hypothesis A:
Hypothesis B:
Non-Independence
Non-Independence
cost of 7,2%
cost of 14,4%
PIBMIN
-4.3
2.9
PIBINT
27.8
20.6
PIBMAX
57.6
50.4
Panel B
First year of net benefit appearance
PIBMIN
1829
1827
PIBINT
1845
1834
PIBMAX
Post 1860
Post 1860
Source: Dobado y Marrero (2001).
28